Statistical interpretation of mtDNA matches in two uncommon types of forensic cases

Mitochondrial DNA (mtDNA) analysis is a frequently used tool for determining the potential origin of biological traces found at crime scenes. The method typically involves comparing the genetic profile of the trace with that of a suspect. While a mismatch between the two profiles usually leads to the exclusion of the suspect, the evidential value of a match is sometimes difficult to grasp. This is particularly true in cases that are more complex than a simple trace-suspect comparison. We considered two such scenarios and developed means for appropriate statistical interpretation of the respective mtDNA results. One scenario requires the evaluation of a composite hypothesis about trace donorship in multiple cases involving an mtDNA profile match with one and the same suspect. The other scenario calls for the consideration of a second mtDNA profile found at the crime scene that matches a matrilineally unrelated contact person of the suspect. For both scenarios, we propose formally linked mathematical methods for interpreting the mtDNA data which, under certain assumptions, allow valid quantification of the evidential value of the latter for or against the suspect. Furthermore, we illustrate the application of both methods with example calculations under realistic assumptions about the required parameters.